Tesla Releases FSD Safety Data From the Netherlands: 3.5x Fewer Crashes Than Human Drivers
5 min read read
Two months after Full Self-Driving (Supervised) launched in the Netherlands, Tesla has published its first official safety report from European roads — and the numbers are striking. Over the roughly two-month window from April 10 to June 5, 2026, vehicles running FSD in the Netherlands logged 23.6 million kilometres (14.6 million miles) of data. The conclusion: FSD triggered 3.5 times fewer collisions than human drivers covering the same road types.
The report was not voluntary marketing. The Dutch vehicle authority, RDW, required Tesla to commit to annual performance reporting as a formal condition of the April 10 type approval. This makes the Netherlands dataset the first regulatory-mandated FSD safety disclosure in any market.
The Numbers by Road Type
| Road Type | FSD Distance | FSD Collisions | Manual Distance | Manual Collisions | Safety Ratio |
|---|---|---|---|---|---|
| Highways | 16.6M km | 0 | 158.7M km | 33 | FSD: zero crashes vs manual baseline |
| Non-highway roads | 7.0M km | 3 | 152.9M km | 109 | 1.6x safer |
| All roads combined | 23.6M km | 3 | 311.6M km | 142 | 3.5x safer |
On Dutch highways specifically, Tesla vehicles drove 16.6 million kilometres on FSD without a single recorded collision. The manual driving baseline for the same road type produced 33 collisions over 158.7 million kilometres — a collision rate of roughly 1 per 4.8 million km of manual highway driving.
Beyond Collision Counts: Driver Behavior Metrics
The report also captured broader driver behavior data that goes beyond raw crash statistics:
| Metric | FSD vs Manual |
|---|---|
| Automatic Emergency Braking events | 14.9x fewer on FSD |
| Harsh acceleration events | 8.8x fewer on FSD |
| Harsh braking events | 7.3x fewer on FSD |
AEB activations — where the car brakes autonomously to avoid a collision — are a leading indicator of near-misses. The 14.9x reduction in AEB events suggests FSD not only crashes less often, but encounters fewer dangerous situations overall.
Why This Report Carries More Weight Than Tesla's US Data
Tesla publishes quarterly FSD safety reports for the US market, which have consistently shown fewer miles-per-incident than NHTSA's manual driving baseline. Critics have long questioned those reports on methodological grounds — Tesla defines its own metrics, selects its own comparison baselines, and publishes without third-party audit.
This is the first time FSD safety data has been produced under a regulatory compliance obligation rather than as voluntary corporate disclosure. RDW required annual reporting as a condition of the type approval — making this dataset the closest thing to independently mandated safety evidence Tesla has published anywhere in the world. — Not a Tesla App, June 9, 2026
The Netherlands dataset is still relatively small — two months and 23.6 million km — compared to Tesla's US dataset of billions of miles. But the regulatory framing changes its evidentiary standing in the European approval process.
What the Data Does Not Settle
The Netherlands roads are among the most well-maintained and clearly marked in Europe. Dutch drivers are accustomed to predictable traffic patterns, and weather in the April–June window was mild. Critics will note that the same performance cannot be assumed for icy Scandinavian roads, complex German autobahns with unlimited-speed sections, or dense urban driving in Paris or Rome.
Denmark's Road Traffic Authority, which just provisionally approved FSD on June 9, had specifically cited icy road performance as one of its earlier concerns. The Dutch dataset covers the best-case European conditions, not the worst.
The Bottom Line for the EU Approval Debate
The Netherlands safety report gives Tesla its strongest European regulatory argument to date. With a 3.5x overall safety improvement and zero highway crashes across 16.6 million kilometres, the data will be central to any EU Commission review. The next anticipated vote on bloc-wide FSD approval is estimated for October or December 2026 — and this report is the most concrete safety evidence available ahead of that decision.
Photo: Tesla Model 3 urban driving / Pexels